Energy efficiency and environmental considerations for green data centres

144

Int. J. Green Economics, Vol. 8, No. 2, 2014

Energy efficiency and environmental considerations for green data centres Mueen Uddin* Faculty of Computer Systems & Software Engineering, Universiti Malaysia Pahang, Kuantan, Pahang, Malaysia Email: [email protected] *Corresponding author

Asadullah Shah Department of Information Systems, Kulliyyah of Information Communication Technology, International Islamic University Malaysia, Selangor, Malaysia Email: [email protected]

Jamshed Memon Department of Information Systems, Faculty of Computing, Universiti Teknologi Malaysia, Skudai, Malaysia Email: [email protected] Abstract: The advancement in information and communication technology based businesses and social practices in last few decades have transformed many, if not most, economies into e-economy and businesses into e-businesses. Technology has the potential to create sustainable business and society both in grim and green economic times. Data centres are found major culprits in consuming too much energy in their overall operations and generating huge amount of CO2. This paper determines the properties and attributes of green IT infrastructures and provides ways for achieving green sustainable businesses. The proposed green IT attributes and characteristics using virtualisation technology are very industrious and efficient for data centres to be more energy efficient and green, hence reducing the emission of greenhouse gases so that the overall effect on global warming can be reduced or even eliminated. Keywords: energy efficiency; green IT attributes; green IT; environmental sustainability. Reference to this paper should be made as follows: Uddin, M., Shah, A. and Memon, J. (2014) ‘Energy efficiency and environmental considerations for green data centres’, Int. J. Green Economics, Vol. 8, No. 2, pp.144–157.

Copyright © 2014 Inderscience Enterprises Ltd.

Energy efficiency and environmental considerations for green data centres 145 Biographical notes: Mueen Uddin is Senior Lecturer at Faculty of Computer Systems and Software Engineering, University Malaysia Pahang. He completed his PhD from Universiti Teknologi Malaysia (UTM) in 2013 and MS in Computer Science from Isra University Hyderabad Pakistan with specialisation in information networks. His research interests include green computing, energy efficient data centres, green metrics, global warming effects, virtualisation and cloud computing. He has published more than 65 international papers in reputed indexed journals. Asadullah Shah is a Professor at International Islamic University Malaysia. He earned his PhD in Multimedia Systems from the University of Surrey, UK. His research interests include information system management, multimedia media systems, and animation and compression techniques. He is a member of the IEEE and ACM. He is a renowned researcher with over 80 publications in journals and conferences. Jamshed Memon is PhD fellow at Universiti Teknologi Malaysia. He completed his MS in Computer Sciences from Greenwich University, UK, in 2009. His research interests include entrepreneur ecosystems, network security, intrusion detection and prevention systems, green IT, virtualisation and ecofriendly technologies. He is a member of the IEEE and AIS. He has published more than ten international journal papers in indexed journals.

1

Introduction

Seldom does a day pass in which we do not hear or read about sustainability or ‘going green’. Environmental concerns are constantly in news headlines, and the impact of technology on our environment is significant. Large technology organisations such as Dell, HP, ISM, Sun, Hitachi, and Fujitsu have introduced green and sustainable initiatives. ‘Green’ is generally understood to mean ‘friendly to the environment and energy efficient’. Sustainable implies planning and investing in a technology infrastructure that serves the needs of today as well as the needs of tomorrow while conserving resources and saving money. Organisations are quite concerned with environmental issues, but they have also come to realise that sustainable business practices can significantly enhance the bottom line. The data centres have become an increasingly important part of most business operations in the 21st century. With escalating demands and rising energy prices, it is essential for the owners and operators of these mission critical facilities to assess and improve their performance. In contexts ranging from large-scale data centres to mobile devices, energy use is an important concern. In data centres, power consumption in the USA has doubled between 2000 and 2006, and will double again in the next five years (EPA, 2006). Server power consumption not only directly affects a data centre’s energy costs, but also necessitates the purchase and operation of cooling equipment, which can consume one-half to one watt for every watt of power consumed by the computing equipment (Rivoire, 2008). As new servers are being added continuously into data centres without considering the proper utilisation of already installed servers, it will cause an unwanted and unavoidable increase in the energy consumption, as well as increase in physical infrastructure like over-sizing of heating and cooling equipments. This increased consumption of energy causes an increase in the production of greenhouse

146

M. Uddin, A. Shah and J. Memon

gases that are hazardous for environmental health. Hence it not only consumes space and energy but also cost environmental stewardship (Uddin and Rahman, 2012). The continued growth of data centre power consumption impacts everything from the business enterprise to the power supply companies to the environment. With more efficient energy use in data centres, power supply companies will face less demand and the possibility of excess power, which could help limit blackouts, reduce carbon dioxide output and cut other greenhouse gases. In addition, energy use has implications for reliability, density and scalability. As data centres house more servers and consume more energy, removing heat from the data centre becomes increasingly difficult (Chandrakant, 2003). Since the reliability of servers and disks decreases at high temperatures, the power consumption of servers and other components limits the achievable density of data centres, which in turn limits their scalability. Furthermore, energy use in data centres is starting to prompt environmental concerns of pollution and excessive load placed on local utilities (Chandrakant and Ranganathan, 2006). These concerns are sufficiently severe that large companies are starting to build data centres near electric plants in coldweather environments (Markoff and Hansell, 2006). For the business enterprise, an increase in data centre efficiency can save significant energy costs. However, even with the global presence of many companies, these metrics are often not applied consistently at a global level. All of these factors are increasing the public’s awareness and global concerns of these current power consumption trends. We are in the biggest data centre construction boom in history. The USA is spending $16 billion a year building out additional data centres, with another $6 billion on refurbishing existing ones. Experts say we should be spending $3 billion a year to build new electrical power plants to meet the supply needs of these data centres – except we are not (EPA, 2009). The Smart 2020 report published by the Climate Group and GeSI revealed that in 2002 the global data centre footprint, including equipment use and embodied carbon, was 76 MtCO2e and this is expected to more than triple by 2020 to 259 MtCO2e, making it the fastest-growing contributor to the Information and Communication Technologies (ICT) sector’s carbon footprint, at 7% per annum in relative terms. If growth continues in line with demand, the world will be using 122 million servers in 2020, up from 18 million today (Clarke et al., 2006). With energy prices increasing worldwide the operational costs of data centre continue to increase steadily. Besides the cost, availability of electrical power is becoming a critical issue for many companies whose data centres have expanded steadily. Enterprises, governments and societies at large have a new important agenda: tackling environmental issues and adopting environmentally sound practices. Over the years, the use of IT has exploded in several areas, improving our lives and work and offering convenience along with several other benefits. We are passionate about advances in and widespread adoption of IT. However, IT has been contributing to environmental problems, which most people do not realise. Computers and other IT infrastructure consume significant amounts of electricity, placing a heavy burden on our electric grids and contributing to greenhouse gas emissions. Additionally, IT hardware poses severe environmental problems both during its production and its disposal. IT is a significant and growing part of the environmental problems we face today. We are obliged to minimise or eliminate where possible the environmental impact of IT to help create a more sustainable environment. To reduce IT’s environmental problems and to create a sustainable environment, we call upon the IT sector as well as every computer user to green their IT systems, as well as the way they use these systems (Uddin et al., 2014).

Energy efficiency and environmental considerations for green data centres 147 We are legally, ethically and socially required to green our IT products, applications, services and practices. Green IT benefits the environment by improving energy efficiency, lowering greenhouse gas emissions, using less harmful materials and encouraging reuse and recycling. Factors such as environmental legislation, the rising cost of waste disposal, corporate images and public perception give further impetus to the green IT initiative. Green IT is a hot topic today and will continue to be an important issue for several years to come. To foster green IT, we should understand: what are the key environmental impacts arising from IT? What are the major environmental IT issues that we must address? How can we make our IT infrastructure, products, services, operations, applications and practices environmentally sound? What are the regulations or standards with which we need to comply? How can IT assist businesses and society at large in their efforts to improve our environmental sustainability? This paper highlights some of these issues, and then presents a holistic approach to greening IT in e-businesses especially data centre industry. We propose a green IT strategy for data centres and outline specific ways to minimise IT’s environmental impact.

2

Environmental impact of IT

For over a quarter century, researchers have been cognisant of the potential for the rise in trade to negatively impact the environment. There is a global change in the world’s climate due to intensifying emissions of greenhouse gases and other fuel combustions, causing potentially disastrous consequences, and as these problems become global, a new spotlight appears on IT. IT is turning out to be both a solution and problem for environmental sustainability. Unfortunately, it is contributing enormously towards environmental problems causing gigantic problem of global warming. Its immense use has exploded in almost all areas of business activities offering great benefits and convenience and at the same time transforming businesses and societies into global world (Molla, 2008). IT affects our environment in several different ways. Each stage of a computer’s life from production, use to disposal presents environmental challenges. Manufacturing computers and their various electronic and non-electronic components consume electricity, raw materials, chemicals, water and generate hazardous waste. The increased number of computers in use and their frequent replacement make the environmental impact of IT a major concern. This increase in energy consumption results in increased emission of greenhouse gases as most of the electricity is generated by burning coal, oil or gas. All these factors contribute towards environmental problems. As businesses balance their growth and production with environmental risks, they are baptised to innovate and implement greener solutions to make IT systems and work practices energy efficient and environment friendly (Fuchs, 2008). Data centre operational managers need to implement green IT techniques and policies to reduce greenhouse gas emissions by improving energy efficiency, thereby reducing the rate of energy consumption by improving the utilisation of already installed equipments (Petrucci et al., 2009). Other policies, like the use of lower emission, clean fuels such as wind, solar, biomass or hydroelectric would reduce total emissions by reducing emission intensity while maintaining the rate of energy consumption (Bhutto and Karim, 2007). Today global warming has been a much-debated issue amongst the scientific society. The consensus amongst most of the scientific community is that the problem of global warming is largely due to an increase in greenhouse gasses in the atmosphere as a result

148

M. Uddin, A. Shah and J. Memon

of human activity (Packard and Reinhardt, 2007). Energy savings play a key role in reducing the overall cost of ownership of data centre, and is being discussed everywhere for implementing and achieving businesses with sustainable development considering economic, social and environmental dimensions. One of the ways the organisations can reduce their carbon footprints and lessen their harmful environmental impact is through adopting environmentally friendly Information Technology (IT) practices, frequently referred to as ‘green IT’ (Tracy et al., 2011). These practices mitigate environmental risks associated with the emergence of a green economy. Organisations that mitigate these risks more effectively than their competitors would be able to gain a competitive advantage. Global warming is a matter of high concern for all inhabitants living in this planet. Individuals, organisations and governments are doing their best to control the warming and make this tiny planet a liveable place. Information technology, telecom/data network can play an equally important role along the side of other direct means for solving this menace. Many technologies are in place and others are to be developed or fine tuned to make them ready for use in reducing the emission of greenhouse gases into atmosphere (Jacobson, 2009). It has taken more than 30 years for data centre industry to evolve to a point of modularity that is conducive to alignment of efficiencies in use and design across disparate infrastructures. The green movement has been around for many years to evolve to a point where the economics of sustainable practices are well understood. Traditionally, environmentalism has been perceived to be at odds with economic prosperity. Environmental stewardship encompasses the notion of balancing current resource consumption with the resource requirements of future generations. The landmark Brundtland report defines sustainable development as meeting “the needs of the present without compromising the ability of future generations to meet their own needs” (Benjamin and Marilyn, 2010). Green IT extents many focus areas and activities, including power management; data centre design, layout, and location; the use of biodegradable materials; regulatory compliance; green metrics and green labelling; carbon footprint assessment tools and methodologies; and environment-related risk mitigation. A growing number of IT vendors and users have begun to turn their attention towards green IT triggered by the imminent introduction of more green taxes and regulations; there will be a major increase in demand for green IT products and solutions. Green IT is becoming a hot topic these days and for years to come, as it becomes imperative to develop environmentally sustainable IT, from both economic and environmental viewpoint (Murugesan, 2008).

3

Challenges faced by data centres

There is no single bad guy that can be blamed for IT inefficiency. Worse, inefficiency seems to grow incrementally over time as environment becomes older and more complex. Each new application being added seems to require another server, which requires administrative time to keep running, while it consumes power, space and expensive network ports in your data centre. The data centre industry has a number of related problems as following:

Energy efficiency and environmental considerations for green data centres 149 1

Inconsistent measuring metrics and benchmarking: It is very greatly important for data centre managers to measure the performance of their data centre regularly so that efficiency measures should be performed to make data centres energy efficient and green. But unfortunately there is no industry standard metric available acceptable worldwide to measure the performance in terms of energy efficiency and CO2 emissions. Data centre managers are currently equally split between using external benchmarks, home grown tools, financial analysis and commercial asset/financial management tools, with no clear leader and metric. It is evident from different discussions that measuring IT performance is difficult.

2

Large number of underutilised servers: Servers are the major components responsible for performing most of the processing being performed in data centres. There number is continuously increasing as the demands from businesses grow. Owing to their increased number they are the leading consumer of IT power in any data centre. Data centres are plagued with thousands of the server’s mostly underutilised, having utilisation ratio of only 5–10% consuming huge energy and generating huge amount of greenhouse gases (Uddin and Rahman, 2012).

3

Power efficiency of IT equipment: Data centre comprises many types of equipment such as servers, UPS, PDUs, chillers, Cracks. All of these components consume enormous amount of power to provide services to end users. Most of the data managers think that IT equipments are significant source of electrical waste. Proper efficiency measures can reduce these consumptions and help data centre managers implement environment friendly and green data centres.

4

Establishing performance requirements and maximising IT operations: Effective application service delivery requires a continuous understanding of end-to-end application performance requirements. In a data centre environment, with rapidly changing dynamic workload and resource allocation, continuous measurement to establish performance requirements is especially vital. This understanding should start when the applications are still in development, so that IT can avoid any surprise performance problems during and immediately after production deployment. As application usage changes, continuous measurement is required to adapt workload and resource allocation and maintain service levels. When application changes are made, or new features are added, performance requirements will need to be reestablished to again avoid potential disruption.

5

Environmental issues and problems: The growing accumulation of greenhouse gases is changing the world’s climate and weather patterns, creating droughts in some countries and floods in others. It is slowly pushing global temperatures higher, posing serious problems to the world. For instance, 2005 was the warmest year on record, and the ten warmest years have all occurred since 1980. Global data show that storms, droughts and other weather-related disasters are growing more severe and more frequent. To stop the accumulation of greenhouse gases in the atmosphere, global emissions would have to stop growing. Electricity is a major cause of climate change, because the coal or oil that helps generate electricity also releases carbon dioxide, pollutants and sulphur into the atmosphere. These emissions can cause respiratory disease, smog, acid rain and global climate change. Reducing electric power consumption is a key to reducing carbon dioxide emissions and their impact

150

M. Uddin, A. Shah and J. Memon on our environment and global warming. With this in mind, let us focus on what each of us – as IT professionals, members of the IT industry, and IT users – can do individually and collectively to create a sustainable environment. Let us examine IT’s environmental impact and consider green IT measures that we can adopt.

4

Proposed work

The continued rise of internet and web applications is driving the rapid growth of data centres. Enterprises are installing more servers or expanding their capacity. The number of server computers in data centres has increased sixfold to 30 million in the last decade, and each server draws far more electricity than earlier models. Aggregate electricity use for servers doubled between 2000 and 2005, most of which came from businesses installing large numbers of new servers (Pritchard, 2007). One problem with the greening of IT is that it forces organisations to buy more. Plans usually call for things like more energy efficient servers, intelligent sensors for data centre cooling, server virtualisation software, low power monitors and devices that turn off dormant computers (Baines, 2007). The social, financial and practical constraints involved will force businesses and IT departments to reduce energy consumption by data centres. We can improve data centre efficiency by using new energy-efficient equipment, improving airflow management to reduce cooling requirements, investing in energy management software, and adopting environmentally friendly designs for data centres and new measures to curb data centres energy consumption. This paper highlights the importance of green IT for data centres, and proposes attributes that provide data centre managers with guidelines and steps to be followed to make data centres energy efficient and green.

4.1 Proposed energy efficiency attributes for greening data centres The proposed energy efficient green IT attributes help to achieve energy consumptions, underutilisation, emission of greenhouse gases, environmental concerns, global warming issues and intensive administrative labour that contribute towards data centre inefficiency. The proposed attributes highlight top IT improvements in data centre spanned across data centre energy efficiency, infrastructure consolidation, reduced administrative labour, better IT process, improved service time and reduced greenhouse gases to reduce the effects of global warming hazardous for environmental health. These benefits may vary significantly across businesses of different types. In particular, companies with less than $1 billion in revenue said that they benefited more from physical consolidation of IT assets, whereas companies with $5 billion or more in revenue benefited the most from improving the energy efficiency of their data centres. These attributes are as follow: 1

Baseline your environment

2

Virtualise servers

3

Consolidate IT

4

Improve data centre efficiency

5

Update IT processes.

Energy efficiency and environmental considerations for green data centres 151

4.1.1 Calculate baseline values The first step in greening the data centre is to baseline all the requirements to get the maximum value out of data centre greening programme. Now more than ever, energy efficiency seems to be on everyone’s minds. Faced with concerns such as global warming and skyrocketing energy costs, more and more companies are considering if and how to increase efficiency. E-businesses that rely on data centres must make hard decisions to accommodate growing demands without creating a negative impact on their finances or the environment. The data centre baseline study report must be based on indepth interviews with engineers and data centre managers. These professionals represent a cross-section of companies in terms of industry, size, number of servers, storage capacity, age, geography etc. The baseline study provides measures to boost efficiency, as well as the incentives for making changes data centre energy efficient. The growth in IT demand is among the most common obstacles to becoming more energy efficient. Data centres must contend with constant expansion in data volume, along with new and extended application requirements. The study also gets awareness about how to calculate the power load of individual IT devices. The data centre baseline report also outlines helpful strategies for approaching energy efficiency in the data centre. The discussion covers virtualisation, airflow management, server decommissioning, equipment upgrades, storage consolidation and optimisation, and use of fresh-air cooling and renewable energy sources. The report also includes tips on how to improve energy efficiency in the data centre so that other data centre professionals can evaluate their options and identify the most appropriate steps for their particular organisations. The process of creating the baseline of your data centre starts by creating an inventory of all resources including servers, resources they require, available resources and their associated workloads, this process is called discovery process. The inventory process includes both utilised and idle servers. It also includes information related to (Uddin and Rahman, 2011).  Make and model of the processor  Types of processors (socket, core, threads, cache)  Memory size and speed  Network type (number of ports, speed of each port)  Local storage (number of disk drives, capacity, RAID)  Operating system and their patch levels (service levels)  Applications installed  Running services. Inventory process It is very important for an organisation to know in advance the total content of its infrastructure before implementing green IT techniques. This is the most important step in greening IT project. There are many tools available from different vendors for performing initial analysis of an organisation. Microsoft Baseline Security Analyser (MBSA) tool provides different information such as IP addressing, operating system,

152

M. Uddin, A. Shah and J. Memon

installed applications and most importantly vulnerabilities of every scanned system. After analysing, all generated values are linked to MS Visio, which generates a complete inventory diagram of all components and also provides details about each component being analysed (Merkel et al., 2010). Microsoft Assessment and Planning (MAP) toolkit is another tool for the assessment of network resources. It works with Windows Management Instrumentation (WMI), the remote registry service or with simple network management protocol to identify systems on network (Koomey, 2007). VMware, the founder of X-86 virtualisation, also offers different tools for the assessment of servers that could be transformed into virtual machines. VMware Guided Consolidation (VGC) a powerful tool assesses network with fewer than 100 physical servers. Since VGC is an agent less tool it does not add any overhead over production server’s workload (Uddin and Rahman, 2011). Categorise server resources After creating server inventory information, the next step is to categorise the servers and their associated resources and workloads into resource pools. This process is performed to avoid any technical political, security, privacy and regulatory concern between servers, which prevent them from sharing resources. Once analysis is performed, we can categorise each server roles into groups. Server roles are categorised into following service types:  Network infrastructure servers  Identity management servers  Terminal servers  File and print servers  Application servers  Dedicated web servers  Collaboration servers  Web servers  Database servers. Aggregate utilisation data helps initially to target particular servers and storage devices as candidates for consolidation. But it does not tell the whole story, since many servers are busy for short periods of time on a periodic basis. In that case, system management tools should be used to collect trends for the entire cycle of systems with applications that run on a weekly, monthly or quarterly basis. Some capacity or consolidation planning tools can simplify this task by superimposing historic data for multiple systems to simplify analysis. To optimise the consolidation scenarios consider using performance modelling and consolidation planning tools to analyse different consolidation and virtualisation scenarios. Different consolidation strategies can be followed to consolidate different components of data centre depending on the workloads and nature of consolidation. Modern planning tools take mathematical approach to optimising capacity, while minimising configuration conflicts between dissimilar systems or those that will compete for resources at the same time.

Energy efficiency and environmental considerations for green data centres 153

4.1.2 Virtualisation process Virtualisation promises to dramatically change how data centres operate by breaking the bond between physical servers and the resource shares granted to customers. Virtualisation can be used to ‘slice’ a single physical host into one or more virtual machines (VMs) that share its resources. This can be useful in a hosting environment where customers or applications do not need the full power of a single server. In such a case, virtualisation provides an easy way to isolate and partition server resources. The abstraction layer between the VM and its physical host also allows for greater control over resource management. The CPU and memory allocated to a virtual machine can be dynamically adjusted and live migration techniques allow VMs to be transparently moved between physical hosts without impacting any running applications. Server virtualisation has become popular in data centres since it provides an easy mechanism to cleanly partition physical resources, allowing multiple applications to run in isolation on a single server. It categorises volume servers into different resource pools depending on the workloads they perform, and then server consolidation is applied. This technique decouples software’s from hardware and splits multi-processor servers into more independent virtual hosts for better utilisation of the hardware resources, allowing services to be distributed one per processor. In server consolidation, many small physical servers are replaced by one large physical server to increase the utilisation of expensive hardware resources, reducing the consumption of energy and emission of CO2 (Uddin and Rahman, 2012). Server virtualisation complements overall IT consolidation projects by allowing firms to share capacity across multiple underutilised systems and shrink the hardware footprint of applications that cannot be completely eliminated. Data centre managers should focus on reducing hardware and operational costs with virtual servers, yet overlook significant improvements to disaster recovery and faster time to market for applications. By offering improved service levels for virtualised servers, we can accelerate internal customers’ migration to virtual infrastructure, while improving overall satisfaction with IT services. Physical to virtual live migration This is the most critical, time-consuming and painful operation when performed manually, since it includes cloning existing operating system and restoring it on an identical machine, but at the same time changing the whole underlying hardware, which can lead to driver reinstallation or possibly the dreadful blue screen of death. To avoid these ambiguities, virtualisation vendors started to offer different physical to virtual (P2V) migration utilities. This utility software speeds up the movement of operation and solves on the fly driver incompatibilities, by removing physical hardware dependencies from server operating systems and allowing them to be moved and recovered. Instead of having to perform scheduled hardware maintenance at some obscure hour over the weekend, server administrators can now live migrate a VM to another physical resource and perform physical server hardware maintenance in the middle of the business day. Virtuozzo for Windows 3.5.1 SWsoft itself introduced a Physical to Virtual (P2V) migration tool called VZP2V. This tool can remotely install P2V knowing machine administrative username and password.

154

M. Uddin, A. Shah and J. Memon

Management to increase utilisation ratio Server consolidation increases the utilisation ratio of underutilised volume servers from 10% to 50% or even more by proper management of workloads to be virtualised to increase the productivity of data centre and reduces the total cost of ownership. There is always a room for improvement, however, as many data centres leave a substantial amount of headroom on their virtual server hosts. Today, some data centres are consolidation the load of 5–10 virtual machines (VMs) on single server, while more experienced organisations are putting 25–30 VMs on a single server. Many administrators are reluctant to run servers at maximum capacity because they are concerned about the possibility of performance problems that could affect multiple applications simultaneously. In order to get to higher levels of hardware utilisation, it is important to improve the administrators’ visibility into the performance and availability of the virtual infrastructure with management tools designed for virtual servers. Active power management software can be used to help power your server infrastructure up and down depending on the demand for applications. This is particularly useful in virtual environments where live migration is used to consolidate VMs onto as few physical servers as necessary to maintain service levels, shutting down the rest.

4.1.3 Consolidate IT resources The best way to reduce hardware, software, labour and facilities costs is to unplug unneeded infrastructures. But it is a complex task that requires a lot of legwork and detailed information on your assets to do it right. You need to assess data centre from all aspects and then categorise it into measurable units so that consolidation can be applied and then benchmarking can be set properly to reduce the consumption of energy and emission of greenhouse gases. IT consolidation involves the consolidation of servers, storage devices, applications running and operating systems. Virtualisation increases the capability of already installed equipments by increasing their utilisation ration and thus reduces the overall operational costs. Many data centre managers tend measure the consolidation success rate by the percentage reduction in their IT budget, while it is important to note that consolidation success rates should be measured from percentage reduction in operational costs. Virtualisation complements IT consolidation but cannot replace it. Even after virtualising, you will still be paying for the maintenance of the same number of application instances, even if they use less equipment to run. As a result, companies frequently struggle to reduce operational costs on the basis of virtualisation alone. Direct-attached storage is usually blamed for data centres low storage utilisation. However, networked storage can also suffer because of over-provisioning and isolated Storage Area Networks (SANs). You may already be paying for intelligent arrays with virtualisation, thin provisioning or reduplications features waiting to be turned on.

4.1.4 Improve data centre efficiency Considering the power consumption in data centres, the main problem is the minimisation of the peak power required to feed a completely utilised system. In contrast, the energy consumption is defined by the average power consumption over a period of time. Therefore, the actual energy consumption by a data centre does not affect the cost of the infrastructure. On the other hand, it is reflected in the electricity cost consumed by the system during the period of operation, which is the main component of a data centre’s

Energy efficiency and environmental considerations for green data centres 155 operating costs. Furthermore, in most data centres 50% of consumed energy never reached the computing resources: it is consumed by the cooling facilities or dissipated in conversions within the UPS and PDU systems. With the current tendency of continuously growing energy consumption and costs associated with it, the point when operating costs exceed the cost of computing resources themselves in few years can be reached soon. Therefore, it is crucial to develop and apply energy-efficient resource management strategies in data centres. Upgrades the data centre power and cooling infrastructures, so that energy efficiency can be achieved. There are many opportunities to reclaim capacity (and reduce electrical costs) in data centre by making both small and large adjustments. Improving data centre efficiency is especially important for large data centres. Prevent hot and cold air mixing When hot exhaust air mixes with cold air, it increases the intake temperature of equipments installed in the data centre. This means that it is necessary to set the temperatures even lower, to accept the intake temperatures. It is necessary to isolate the exhaust air with a hot aisle containment system or ceiling to reduce the load on cooling system and increase the power density of racks. Because it is relatively inexpensive, compared with new infrastructure, upgrade facilities to prevent hot/cold air mixing were one of the top choices among the data centres. Before overhauling anything, start small improvements like eliminating under-floor obstructions to airflow, plugging cable cut outs, and installing blanking panels in racks can improve the amount of air delivered to racks. Older, uninterruptable power supplies and power distribution units may have older, less efficient transformers that are responsible for a sizable portion of the wasted electricity in data centre. It is important for data centres to replace these older systems with newer, more efficient. Most data centre infrastructures are network oriented allowing collecting usage statistics from a variety of energy management software.

4.1.5 Update IT processes Consolidation and virtualisation helps to optimise hardware and software investments, reduce the number of systems being managed, and free up or close some underutilised volume servers. However, IT processes remain unchanged and probably a major source of IT inefficiency. Many data centres are implementing more formalised IT processes, while others suffer from IT processes that have too many steps and depends on manual labour to get them done. Therefore, it now becomes obvious for data centre managers to revise their IT processes to achieve green data centres. The focus should be on those processes that are most critical to running reliable and efficient IT services. In particular, problem management and incident management issues in data centre followed by financial management and configuration management issues. IT processes ensures more reliable services, but with added records and data formalised processes do not get better efficiency because of the added overhead. The efficiency can be achieved by upgrading the system management tools that integrate with service desk software and provide more task-level automation to free up administrators’ time.

156

5

M. Uddin, A. Shah and J. Memon

Conclusion

Green IT is constantly becoming more relevant, and many organisations are working towards reducing the carbon footprint of their data centres. This reduction in carbon footprint is achieved by reducing the data centres power consumption, which in turn results in savings for the organisation. Many new techniques have been used to achieve this reduction in power. One of them is virtualisation. It helps to consolidate multiple servers onto a few physical machines, which increases their utilisation and decreases their power consumption. This paper proposes green IT attributes as solution for implementing green data centres. The proposed attributes provide a solution based on virtualisation technology to overcome the issues and challenges of data centres like energy efficiency and CO2 emissions to reduce the effects of global warming.

References Baines, S. (2007) ‘The greening of IT: why less is more’, Silicon.Com, 13 August. Available online at: http://editthis.info/lauraibm/The_greening_of_IT:_Why_less_is_more Benjamin, K.S. and Marilyn, A.B. (2010) ‘Competing dimensions of energy security: an international perspective’, Annual Review of Environment and Resources, Vol. 35, pp.77–108. Bhutto, A.W. and Karim, S. (2007) ‘Energy-poverty alleviation in Pakistan through use of indigenous energy resources’, Energy for Sustainable Development, Vol. 11, No. 1, pp.58–67. Chandrakant, D.P. (2003) ‘A vision of energy-aware computing from chips to data centers’, Proceedings of the International Symposium on Micro-Mechanical Engineering (ISMME), 1– 3 December, Tsuchiura. Chandrakant, D.P. and Ranganathan, P. (2006) ‘Enterprise power and cooling’, Tutorial at the International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), October. Clarke, L., Weyant, J. and Birky, A. (2006) ‘On the sources of technological change: assessing the evidence’, Energy Economics, Vol. 28, pp.579–595. EPA (2006) U.S. Environmental Protection Agency, ENERGY STAR Program. Report to Congress on Server and Data Center Energy Efficiency (online). EPA (2009) EPA ENERGY STAR Program Requirements for Computer Systems – Draft 4, Environmental Protection Agency, Washington, DC. Fuchs, C. (2008) ‘The implications of new information and communication technologies for sustainability’, Environment, Development and Sustainability, Vol. 10, No. 3, pp.291–309. Jacobson, M.Z. (2009) ‘Review of solutions to global warming, air pollution, and energy security’, Energy & Environmental Science, Vol. 2, No. 2, pp.148–173. Koomey, J.G. (2007) Estimating Total Power Consumption by Servers in the US and the World, Lawrence Berkeley National Laboratory, Berkeley, CA. Markoff, J. and Hansell, S. (2006) ‘Hiding in plain sight, Google seeks more power’, New York Times, 14 June. Merkel, A., Stoess, J. and Bellosa, F. (2010) ‘Resource-conscious scheduling for energy efficiency on multicore processors’, Proceedings of the 5th European Conference on Computer systems, ACM, pp.153–166. Molla, A. (2008) ‘GITAM: a model for the adoption of green IT’, 19th Australasian Conference on Information Systems, Christchurch, New Zealand. Murugesan, S. (2008) ‘Harnessing green IT: principles and practices, IT Professional, Vol. 10, No. 1, pp.24–33.

Energy efficiency and environmental considerations for green data centres 157 Packard, K.O. and Reinhardt, F.L. (2007) What Every Executive Needs to Know About Global Warming. In Harvard Business Review on Green Business Strategy, Harvard Business School Press, Boston, MA, pp.21−39. Petrucci, V., Loques, O. and Mossé, D. (2009) ‘A dynamic configuration model for power-efficient virtualized server clusters’, Proceedings of the 11th Brazilian Workshop on Real-Time and Embedded Systems (WTR 2009), Recife, Brazil. Pritchard, S. (2007) ‘IT going green: forces pulling in different directions’, Financial Times, 30 May. Rivoire, S.M. (2008) Models and Metrics for Energy-Efficient Computer Systems, Doctoral Dissertation. Tracy, A., Jenkin. T.A., Webster, J. and McShane, L. (2011) ‘An agenda for ‘Green’ information technology and systems research’, Information and Organization, Vol. 21, No. 1, pp.17–40. Uddin, M. and Rahman, A.A. (2012) ‘Energy efficiency and low carbon enabler green IT framework for data centers considering green metrics’, Renewable and Sustainable Energy Reviews, Vol. 16, pp.4078–4094. Uddin, M., Alsaqour, R., Shah, A. and Saba, T. (2014) ‘Power usage effectiveness metrics to measure the performance and efficiency of data centers’, Applied Mathematics and Information Sciences, Vol. 8, No. 5, pp.2207–2216. Uddin, M. and Rahman, A.A (2011) ‘Techniques to implement green data center to achieve energy efficiency & reduce global warming effects’, International Journal of Global Warming, Vol. 3, No. 4, pp.372–389.